Today was the second day of my 4-day trip to Iowa State for the Atanasoff Symposium.
Warning: The pictures in the LJ-cuts are quite huge at the moment (apologies in advance for the wanton slaughter of bandwidth). I will replace these with more electron-saving versions later. For now, you might not want to click unless you have broadband.
Edit, 13:00 CST Wed 05 Nov 2003: You can click now, the pictures are much smaller.
0715-0805: Up with the sun... Well, kindasorta. I slept suprisingly well, much better than I had expected for my first night in the room where my late grandfather had spent his last days in 1999. After changing, I went downstairs and grabbed the small glass of 2% milk that is my usual breakfast.
0805-0825: It took just 10 minutes to get to the Scheman Building, and after circling around to the front, my folks dropped me off. I took the elevator to the first floor and found the nicely-appointed conference registration desk and Diane Rover (the posters chair) just as a news team was photographing John Atanasoff II. I browsed the table full of books written by the invited speakers and about the subject of the symposium. Then I stashed the nice coffee mug they gave us under a table along with my poster tubes and went up to the second floor.
0825-0835: By this time Vasant Honavar (the Computational Intelligence workshop chair) was busily getting ready for James Hendler's talk on the Semantic Web and OWL. I said hello to him and to Kai-Ming Ho, a colleague of my dad's younger brother in the physics department. Kai-Ming introduced me to Bruce Harmon, a university distinguished professor in physics (like Dr. Ho) and one of the senior scientists at Ames Lab who is working on quantum computing. We chatted very briefly before his session was due to start. Then I dashed downstairs to put up two of my posters.
08:25 CST - This is a replica of the ABC, built from 1994-1997:
Here's a much higher-resolution photo.
0840-0855: Men with tacks... My poster wouldn't stop curling on me, so finally I gave in and re-rolled it in the opposite direction. It still took 16 tacks for the big (4' x 5') poster and 8 for the little (17" x 22") one.
0855-0920: Semantic Web blessed by Tyope! Did you know that the language that James Hendler calls "DAML+OIL v1.0" is spelt OWL because the actual anagram of Web Ontology Language (WOL) denotes Owl in A. A. Milne's Pooh books? Oh, you did? Bother, I didn't. (I'm not sure whether this qualifies as piety or impiety towards Tyope, but in any case, the last half of Hendler's talk was quite informative for me.)
0920-1000: Where's Baldi? Did you know that I'd never actually seen so much as a photo of Pierre Baldi before this morning? Odd. I was even at ISMB-2002 in Edmonton, where my first Ph.D. graduate, Haipeng Guo, was the only one who registered in time for Baldi's standing room only tutorial.
09:30 CST: Pierre Baldi giving a lecture on machine learning applications in genomics and proteomics:
1000-1210: It's all a blur! I spent a bit longer than I planned at my poster - over two hours, to be more precise. I got to talk to lots of Vasant's students and Raghu Ramakrishnan, whom I also hadn't ever met before today. We must have talked for over 20 minutes, and I didn't recognize him until he handed me his business card, even though I met with his former Ph.D. student and co-author Johannes Gehrke, in 1999 when Johannes came to NCSA).
1210-1340: Lunch was fascinating. I sat with several local faculty members and Yuan Lin, a bioinformatics researcher at Integrated DNA Technologies, whose primer design software we used in our NSF Research Experience for Undergraduates (REU) this past summer. It's always cool to meet someone who's responsible for software you use, or who uses your software. (I also met a few people who are on the BNJ developer list.)
12:30 CST: Speaking of NSF, here is the Chief Information Officer (CIO) of NSF and the former department head of Computer Science at Iowa State University, George Strawn, giving the after-luncheon talk:
1340-1430: Not late after all... It turns out that I wasn't late to Michael Pazzani's KDD talk, even though I stayed to the end of George Strawn's talk. That's because Stephen Smale, the second scheduled speaker in the track, couldn't make it. The KDD talk, which outlined several issues and studies on usability of models produced by machine, was quite interesting. It's been a while since I've seen a screen shot from the late, lamented SGI MineSet. As Pazzani pointed out, it hasn't actually been demonstrated (AFAIK) that the "flight simulator" feature actually helps. There were lots of questions from the audience (myself included) about the user study. Personally, I'd like to see more of these - at least, people working with CRM should publish more. Vignettes from Ronny Kohavi et al. are all the typical ML/KDD researcher hears.
13:50 CST: Michael Pazzani of UC Irvine (currently director of the Information and Intelligent Systems division of NSF CISE) giving his leture on KDD:
1430-1510: I talked with Facundo Bromberg and several other grads in the AI group at IA State. robbyjo, Edwin Rodríguez and I had met Facundo at UAI-2003 in Acapulco, Mexico.
1510-1515: The way to a Jedi hobbit's heart is undoubtedly through his stomach, but more specifically, lemon squares are definitely the way to go.
1515-1545: I briefly recapped the history of UIUC's Artificial Neural Networks and Computational Brain Theory (ANNCBT) group for Facundo. Interestingly, a lot of people at KSU-CIS are asking me about ANNCBT-type activities these days. Ironically, Jesse Reichler, who came up with the idea of ANNCBT and led us in founding it, started and quit blogging before it was fully in vogue (1996-2001), but it was his inspiration that got me started blogging.
1545-1620: I caught the tail end of a fascinating talk on wearable computers by Thad Starner, an assistant professor at the Georgia Tech College of Computing.
1620-1705: OK, Raghu Ramakrishnan is a genius, and I don't just mean with respect to databases. :-)
16:20 CST: Now that is some pretty nifty ex tempore PowerPoint trasking:
1705-1805: So, I got to talking with Thad, Daniel Sieworek (a professor at Carnegie Mellon University who directs an HCI lab there), and a fellow from a local company, and it turns out that Thad and Daniel both know my uncle (my mom's younger brother) from his days at CMU-SCS. "Oh, I know Kai-Fu Lee; he taught me LISP", said Thad. Small world, isn't it? We started discussing, in no particular order:
- HCII on wearables as a possible IJCAI-2005 tutorial
- The UIUC Beckman Institute's HCII group
- Mobile speech recognition
- The possibility of putting JVA recordings and slides online and making them freely available by streaming video or webcast cf. Tegrity
- The Illinois/TEPIA teleconference on HCI in November, 1997
- MIT's OpenCourseWare distribution of their recorded lectures
- Operating systems on wearables: (Debian) Linux vs. others
- Hillol Kargupta's work on ubiquitous data mining
- PDAs and the Newton
- Wireless data plans (e.g., Sprint's data phones under the $80 monthly unlimited data plan apparently includes 2500 shared anytime minutes and no limits on the number of phones that can be added), data transfer, and custom tricks (e.g., power over USB)
- Cortical vision implants for the blind
- Virtual keyboards
- Immersadesk/CAVE/C6-type technologies and portable versions thereof
- Ashutosh Garg's demo during Bill Gates's keynote address at keynote talk at IJCAI-2001
- Speech-to-speech translation and the Verbmobil project
- Wolfgang Wahlster's keynote talk at IJCAI-2001
- Haptic feedback devices such as the SensAble Technologies Phantom (a force feedback pen) for teleoperation and telesurgery
Yeah, it was a really interesting hour.
17:10 CST: The ultra-tronkie recipient of Dr.
We started calling ourselves the Borg, because we knew the press would do it sooner or later. - Thad Starner
(Now there is one more tronkie in the world who knows the term. :-))
1805-1925: The banquet dinner was really good. Having it in the Brunnier Art Museum, amidst a collection of early-to-mid-20th century Russian paintings, made me a just a little bit nervous. The dinner conversation continued along the lines of our discussion in the previous hour, and I was quite impressed to learn that Thad not only has 7 Ph.D. students but took them all with him to Switzerland. He gave us the URL of the company that markets some of his wearables... now, tell me honestly: if Holly-Marie Combs (from a certain TV show) were to model one of these ensembles, you'd consider getting one, right? ;-)
I actually took this earlier in the day, but here's the entrance to the museum:
1925-2055: The evening speakers were Iowa State provost Benjamin Allen, John Atanasoff II (son of the late Dr. John Vincent Atanasoff in whose honor the symposium is being held), and Stephen Wolfram, live from Champaign, IL. U.S. Senator Tom Harkin also popped in remotely to say hello and a few words about Atanasoff's life, work, and impact.
20:00 CST: Here's Dr. Stephen Wolfram, developer of Mathematica, giving his talk via videoteleconference (VTC) from his home in chambana:
(I have some screenshots from Wolfram's talk that will appear here later.)
2055-2105 I went out the wrong door! Oops.
This scene, with a hauntingly low moon over the Iowa State campus, looked a lot cooler than it turned out in the picture. I've got to learn how to focus the camera for night shots.
2105-2315: I came back to my grandmother's house and uploaded all the digital photos from my Vision PCS phone. I talked with my grandma for a bit and later with my folks about some of the interesting things I saw today.
Going to bed at 11pm is most unnatural.